Taehee Lee (@tlee_rad) profile picture

Taehee Lee

@tlee_rad

Radiologist, Interested in computational methods in thoracic imaging

github.com/lee8720
Joined May 2023
  • Tweets8
  • Following44
  • Followers2
1Media
Photo by Taehee Lee (@tlee_rad)
Sample size is so very neglected in most prediction model research Here is a paper from our team that guides researchers toward sample sizes for precise performance estimates in validation studies And pmvalsampsize is your one line of code to implement it! By @joie_ensor
The BMJ (@bmj_latest) profile pictureThe BMJ
@bmj_latest27 Jan 2024
How should you calculate the sample size required for external validation studies? @Richard_D_Riley, @GSCollins, and colleagues set out guidance and provide illustrated examples bmj.com/content/384/bmj-2023…
Evaluation of clinical prediction models (part 3): calculating the sample size required for an...

Evaluation of clinical prediction models (part 3): calculating the sample size required for an...

An external validation study evaluates the performance of a prediction model in new data, but many of these studies are too small to provide reliable answers. In the third article of their series on...

bmj.com
1
11
1
60
10,095
Taehee Lee retweeted
The BMJ (@bmj_latest) profile picture
The BMJ
@bmj_latest21 Jan 2024
How should researchers undertake an external validation study? This article guides readers through the steps involved @Richard_D_Riley bmj.com/content/384/bmj-2023…
Evaluation of clinical prediction models (part 2): how to undertake an external validation study

Evaluation of clinical prediction models (part 2): how to undertake an external validation study

External validation studies are an important but often neglected part of prediction model research. In this article, the second in a series on model evaluation, Riley and colleagues explain what an...

bmj.com
9
24
11,944
Very excited to share a new work on AUPRC, AUROC, and class imbalance: arxiv.org/abs/2401.06091 The short of it? Despite widespread belief, AUPRC is not superior to AUROC under general class imbalance and it can exacerbate biases towards subpopulations with higher prevalence.
A Closer Look at AUROC and AUPRC under Class Imbalance

A Closer Look at AUROC and AUPRC under Class Imbalance

In machine learning (ML), a widespread claim is that the area under the precision-recall curve (AUPRC) is a superior metric for model comparison to the area under the receiver operating...

arxiv.org
6
35
9
199
64,509
Reminded today about how important this paper is by @laure_wynants @BenVanCalster @ESteyerberg @MaartenvSmeden et al "Three myths about risk thresholds for prediction models" Highly recommended reading bmcmedicine.biomedcentral.co…

Three myths about risk thresholds for prediction models

BMC Medicine - Clinical prediction models are useful in estimating a patient’s risk of having a certain disease or experiencing an event in the future based on their current characteristics....

link.springer.com
1
21
71
9,116
NEW PAPER in @bmj_latest (with @Richard_D_Riley) - 1st in a 3 part series on the ‘Evaluation of clinical prediction models’. Part 1 is ‘from development to external validation’. —> tinyurl.com/n8fy5xvj
Image from Gary Collins (@GSCollins)
3
90
8
252
79,193
⭐️NEW paper “Clinical prediction models and the multiverse of madness” For any model created, a multiverse of other potential models exists & an individual’s prediction may vary greatly across this multiverse Avengers endgame: examine this instability🙏 bmcmedicine.biomedcentral.co…
Clinical prediction models and the multiverse of madness

Clinical prediction models and the multiverse of madness

BMC Medicine - Each year, thousands of clinical prediction models are developed to make predictions (e.g. estimated risk) to inform individual diagnosis and prognosis in healthcare. However, most...

link.springer.com
5
46
2
141
19,905
What an amazing experience at #RSNA23! Four radiology residents from @SNU_Rad.
Image from Taehee Lee (@tlee_rad)
1
4
286
In a new Perspective, Walia, Tuia & @VPrasadMDMPH discuss the adoption of composite end points in oncology, their pro and cons, and advocate for separate reporting of component events to enable a better understanding of cancer biology and trial results: nature.com/articles/s41571-0…
Image from NatureRevClinOncol (@NatRevClinOncol)
Image from NatureRevClinOncol (@NatRevClinOncol)
Image from NatureRevClinOncol (@NatRevClinOncol)
Image from NatureRevClinOncol (@NatRevClinOncol)
1
48
4
149
229,826
Load more